Fujitsu Unveils Next-Generation FUJITSU-MONAKA 2nm CPU for Sovereign AI Infrastructure
KAWASAKI, Japan — Fujitsu Limited officially announced on September 14, 2026, the global commercial launch of its next-generation microprocessor, FUJITSU-MONAKA, alongside the all-new Fujitsu MONAKA Server designed specifically for sovereign artificial intelligence infrastructure. Built entirely in Japan using an advanced 2-nanometer 3D-stacked architecture, the chip is designed to meet escalating global demands for high-performance enterprise AI processing while drastically cutting data center power consumption.
The FUJITSU-MONAKA CPU operates at a maximum clock frequency of 3.8 GHz with a memory transfer speed of 8,800 MT/s. According to Fujitsu, the processor delivers double the AI inference throughput per server compared to conventional enterprise CPUs while reducing overall power usage by 50 percent. This efficiency allows cloud operators, government agencies, and research institutions to halve the number of physical servers needed to handle intense generative AI workloads.
Global sales of the standalone processor to cloud providers, server original equipment manufacturers (OEMs), and high-performance computing (HPC) centers will commence in November 2026. By focusing on sovereign AI capabilities, Fujitsu aims to provide secure, energy-efficient computing infrastructure for sensitive defense, healthcare, and national security applications in Asia and Europe.
